The goals of this project are to understand the concurrent development of the acoustical cues to sound source location and the neural circuits that encode them. All cues for localization can be captured in measurements of head related directional transfer functions (DTFs), which contain the three main cues, interaural differences in time (ITDs) and level (ILDs) and monaural spectral shape cues. Since the formation of these cues is governed by the linear dimensions of the head and pinnae, the relative magnitudes of the cues will change as these structures grow. These experiments test the hypothesis that during the development of one of the cues, ILD, there is a concomitant development of the ability of cells in the lateral superior olive (LSO), one of the most peripheral sites for binaural interaction, to encode ILDs. Aim I will test the hypothesis that the cues will change in ways predictable from the increasing dimensions of the head and pinnae. DTFs will be measured in kittens aged from 1 week to adult and the cues to location extracted from them. The cues will be compared to measurements of the head and pinnae. Aim II will test the hypothesis that the range of ILDs encoded by LSO cells will increase during development with the expected increase in the magnitudes of the physical ILDs, as measured in Aim I. Also, the ability of cells in the medial nucleus of the trapezoid body, which provide the contralateral inhibitory input to the LSO, to encode the complex spectral shapes necessary for ILD computation will be studied. These studies will determine if and how long these nuclei in the ascending binaural system can compensate for changes in the cues brought about by developmental changes in the acoustical properties of the head and pinnae. Finally, Aim III will study the normal mechanisms by which the cues to location, in particular those based on spectral shape (ILDs and monaural spectral cues), are encoded by the LSO and its afferents in adult cats. Several techniques, including virtual space stimulation and a systems identification method, will be used in Aims II and II1. We hypothesize that across a population of inputs to the LSO, complex spectral shape is represented and that the LSO computes a spectral difference to extract the ILD from these inputs. A better understanding of how complex sound spectra are represented will lead to better designs for the processing strategies in auditory prosthetics like cochlear implants and hearing aids. [unreadable] [unreadable] [unreadable]